In the nature reserve of Avesta, Sweden, a caretaker knelt and aimed to dart one of the three European bison (Bison bonasus, commonly known in Europe as the wisent) that roamed a hundred metres away. He missed. He aimed again. This time the dart hit the bison in the hip. The bison ran a few steps and stood still. His hind legs gave in and, dizzy, he lay down on one side, almost falling, without a sound, eyes open, panting. The bison was tranquillised. A vet ran to him, with a chair in his hand, sat and with swift moves shaved a side of the bison’s shoulder blade. He covered the eyes of the bison with a rag to help him sleep and relax. The vets did not want the bison to know he was being handled, to avoid stress and habituation. The caretakers tied the legs of the bison and dragged the huge limp body, weighing a little more than half a ton, onto a tarp, lifted it onto a platform and then into a truck. The bison would spend a full 26 hours in the confinement of the truck, travelling 3,000 kilometres from Sweden to the South-Western Carpathians of Romania. Sixteen other wisents were being tranquillised and lifted onto trailers on the same day in May 2014, to come together in the Carpathians. The goal of this conservation initiative was to create a thriving free-roaming population in this part of the world where bison had been exterminated by the end of the eighteenth century. Tranquilising, lifting, dragging and transporting are some of the most intensly controlling moments in the relationship between humans and endangered wildlife, stressful and risky interventions mediated by technologies. But these transient heavy interventions are meant to produce less interventionistic relationships in the long run: they are practices leading to reintroducing the animals in areas where they could live independently and roam free.

Starting in 1923, the restoration of the then nearly-extinct wisent was the first international large-scale cooperative project aimed at the preservation of an endangered species in Europe.¹ Hence, despite the species becoming extinct in the wild, the wisents have been bred in captivity and released, leading to the creation of several free-roaming wisent populations in Europe. The first such population was reintroduced in the Białowieża forest in Poland, from 1952. And even though a long time has passed since these beginnings, and the wisent is considered recovered thanks to conservation initiatives,² it is still closely managed by humans. In several places, wisents were until very recently still being fed in winter, managed by culling, even hunted for trophies – rendering controversial the degree of autonomy entailed in the ‘return to freedom’ postulated by reintroductions. The lives of recovered wisent populations are still deeply entangled with humans, with their ideas and practices, with their care and management.

Controversies loom large about which human responses are appropriate in the face of impending extinction. The answers are far from straightforward. Returning endangered species from captivity to autonomy and self-reliance, as the ideas of ‘reintroduction to the wild’ and of ‘rewilding ecosystems’ entail, has become increasingly fashionable in conservation, as opposed to more conventional approaches in which heavy-handed management has been the norm.³ However, even in non-interventionist approaches, tensions have often emerged between how much care and support to give to animals, if any, when to intervene and when to let loose. Conservationists navigate ongoing tensions between intervention and relinquishment, care and containment,⁴ and permanently adjust their practice, dealing in the realm of uncertainty, experimentation and compromise. Non-human autonomy is variable across space and time and should be understood rather as degrees of animal autonomy grounded in histories of human–animal relationships across shared landscapes.⁵ Both intervention and non-intervention can potentially enhance the survival and wellbeing of animals, and at the same time give rise to serious difficulties. They can both open up and foreclose opportunities for flourishing life.

The case of the European bison raises a number of questions: how do conservationists go about their relationships with endangered species? How do they walk the line between intervening and letting loose, between an approach predicated on human responsibility, care and constant support, and a hands-off approach predicated on uncompromising distance and animal agency, promoting self-reliance?

I will explore this tension in the present paper, by looking into the complexities of actual lived practices. I will show how the tension between intervention and non-intervention plays up on two temporal scales. In the long-run, wisent conservation moved slowly from a heavily interventionist, ‘hands-on’ policy of reintroduction, which employed breeding techniques to enhance reproduction, provision of supplementary feed and veterinary care, towards non-intervention, which meant cultivating the autonomy and self-reliance of animals: a move, in short, from an underlying philosophy of ‘breeding’ to a philosophy of ‘rewilding’. However, in the short-run, both ‘hands-on’ and ‘hands-off’ approaches informed practice to different degrees. In the post-war decades as well as in the twenty-first century, both intervention and non-intervention were considered legitimate practices, every situation entailing its own messy complexity. To feed or not to feed; to tranquillise or not to tranquillise; to cull or not to cull – these were (some of) the practical questions.

This paper takes an ethnographic approach integrated in a longer-term historical perspective; a history of reintroductions ‘from below’. I focus primarily on the practices involved in a contemporary reintroduction project in the South-Western Carpathians, which allows for a grassroots view of unfolding human–animal entanglements – the uncertainties and experiences forged in reintroductions on the ground. To gain temporal perspective, I compare this project to the reintroductions of European bison that have occurred since 1952 in Poland – one of the earliest systematic attempts to reintroduce species on the brink of extinction back to the wild (or to a place as wild as a forest in twentieth-century Europe can be). I draw on fieldwork done in 2016 and 2019 in the South-Western Carpthians, interviews, analysis of media sources, scientific reports and articles, as well as secondary literature. I use ethnographic storytelling, and aim to trace the role and agency of the animals themselves in the process.

Scholars warn that we are currently facing a mass extinction event. Yet, daunting as this sounds, recent studies show that conservation has substantially reduced the loss of mammals and birds, and that reintroductions played a major role in this.⁶ Reintroductions are complex practices, and can be largely defined as the return of species that have been lost regionally or globally from the wild, or were critically endangered.⁷ Sometimes this return occurred after years of captive breeding, required to ‘make more’ of animals whose species had diminished to only a few individuals. Reintroductions became popular in the context of the rising plight of endangered wildlife, and of increasing aversion to keeping captive zoo animals, starting in the 1960s and peaking in the middle of the 1980s.⁸ Successful reintroductions involving charismatic mammals – such as the Arabian Oryx in Oman – enhanced awareness about the possibilities of the practice throughout the 1970s and, in the next decade, reintroductions were streamlined through emerging conservation institutions; most notably, in 1988, a reintroduction specialist group was founded within the International Union for Conservation of Nature (IUCN) Species Survival Commission.⁹ Reintroductions gained traction as radical, ‘courageous’ practices, much needed to revolutionise traditional conservation that had achieved little improvement in the face of the desperate plea of shrinking wildlife.

However, reintroductions were chancy endeavours. By the 1960s, ‘precious little was known about the life history and behaviour’ of endangered animals.¹⁰ Data from the 1970s and 1980s suggested that the majority of reintroduction projects failed to establish viable populations.¹¹ The survival of reintroduced animals was challenged and questioned by unexpected incidents and conditions, like harsh winters or disease outbreaks. Scientists and practitioners involved in reintroductions grappled with novelty, uncertainty and ethical dilemmas. Yet, such fissures and ambiguities were creative, allowing reintroductions to unfold as ever-changing processes. ‘[B]y admitting the limits to “knowing the animal”, we allow for the existence of agency in the non-human world’, Karen Jones wrote (of wolves).¹²

The European Bison, because of its affinity to domestic cattle, was somewhat better understood as a species, and in the interwar period breeding techniques and instruments were borrowed from livestock husbandry to manage its recovery, such as the studbook (pedigree books).¹³ But questions emerged as to whether, in fact, such close management would not turn the wisent, surrounded by an aura of romantic wildness, into a domestic creature. Indeed, in reintroductions, close human–animal encounters occured in a contact zone where entwined practices allow for mutual malleability.¹⁴ Reintroductions are deeply relational and situational practices, in which both humans and animals change one another in interaction, the outcomes being contingent upon the finest interactive dynamics. Kept in reserves or zoos, endangered species have often been refashioned as much as rescued by restoration projects.¹⁵ Reintroductions are indeed experiments in the arts of co-existence and -becoming.¹⁶

In the nineteenth century, the European bison was rapidly approaching extinction, but still holding on in its hundreds, notably in game reserves in Poland and in the Caucasus. Valued for its scarcity and wildness, it became a precious possession for museum collections, as well as a gift prized by the European aristocracy,¹⁷ to be displayed in menageries and reserves. However, with the dawn of the twentieth century, things changed. The wild populations of European bison diminished drastically and were ultimately killed off in the 1920s, amid war, firearms, occupations, hunger and confusion. Yet, at the same time, bison was revered by a circle of preservationists, as a noble magnificent animal, a ‘monument’ associated with nobility and primevalness. In 1924, an international plan for preserving the European bison was set in motion.¹⁸ An inventory counted only 54 bison surviving in captivity, which had ‘proven genetic purity’.¹⁹ Abstracted from what was considered their ‘natural’ or ‘true’ spatial context, they were managed intensively in breeding reserves.²⁰ Not all of them reproduced, but eventually twelve did, and became the ancestors of all European bison existing today. As they turned from a local wild animal to an international object of breeding,²¹ European bison numbers grew steadily throughout the decades, and they were eventually set free from enclosures. Nearly thirty years after the initiation of the preservation programme, the first release of European bison back into the Polish Białowieża forest happened in 1952.²² Many other releases into other places followed. To date there are more than 6,000 wisents in 47 free-ranging herds in Europe.²³

In late spring 2014, seventeen European bison arrived by truck in the South-Western Carpathians from zoos and captive-breeding reserves, and made their way out of the trailers after exhausting trips. Looking shabby, with painted spots where they had been injected for tranquilisation, the exhausted animals wobbled onto a straw mat and, regaining their bearings, walked into the fenced area (Figure 1).

The area to which the bison were brought is a mixture of woodlands (some of the last temperate old-growth beech forests of Europe) traversed by creeks and cliffs, interspersed with clearings, grasslands and bordered in the upper part by alpine pastures. Romania has recently become a hotspot for conservation initiatives of all sorts, from curbing illegal logging to creating large private protected areas, restoring degraded landscapes and reintroducing extirpated species, a true conservation frontier.²⁴ The Carpathian Mountains are considered perfect for rewilding bison, less inhabited than the rest of Europe, with high rates of farmland abandonment.²⁵ Yet, the reintroduction area is far from being a no-man’s-land. Human settlements lie not further than a few kilometres from the release area, and a busy motorway passes by. Also, what is good habitat for wisent is not yet settled in science, some ecologists arguing that forests are not ideal, as wisent are essentially steppe grazers that were pushed by human expansion to become refugees in suboptimal forest habitat.²⁶

The first steps of the South-Western Carpathians project happened in a rush. The EU-Life funding scheme allowed for only a short preparation time. The breeders that sent wisents for reintroduction were worried because ‘neither a genetic and demographic management plan, nor a population and habitat viability analysis, were yet in place’ at the time of the first arrival in Romania.²⁷ The project ‘imported’ a large number of animals from all over Europe into an area that did not have any formal protection status; neither were there any staff experienced in conservation on the ground, let alone in bison issues. Romanian specialists and international consultants provided advice; however the people on whose shoulders most of the management work rested, the rangers, were very young locals, who qualified ‘on the job’. Yet, this lack of experience also meant capacity to break new ground, and to attune to the cues given by the animals. The short-term outcomes in terms of numbers were for a while rather inglorious. Initially (in 2014), the project boasted an aim of 500 wisents in the area by 2025,²⁸ which soon proved unrealistic (as of autumn 2021, there are 106 wisent, counting sixteen newborns.)²⁹

Wisents came to the South-Western Carpathians as both a rewilding and a reintroduction project. While both are ‘essentially non-controlling forms of ecological restoration and human interaction with nature’,³⁰ a slight tension exists between them – in rewilding, the focus is on restoring a lost ecosystem, while in reintroduction it is saving a species.³¹ For ‘Rewilding Europe’, the Dutch NGO driving the project, this was a rewilding initiative. As a new paradigm in conservation, rewilding is ‘broadly oriented towards restoring species and, more important, ecological processes that are absent from contemporary landscapes – especially predation, grazing, succession, dispersion, and decomposition’.³² Dutch rewilders mainly promote wildness in their projects, and the idea of ‘letting nature take its course’.³³ Differently from traditional conservation, rewilding involves biopolitics of experimentation rather than control.³⁴ From this perspective, the bison is seen as an agent, a ‘nature engineer’, restoring ecosystems through grazing, browsing, trampling, fertilising soil with manure.³⁵ But the term rewilding is sometimes used to signify reintroductions, the return of captive-born animals into free-range settings³⁶ – rewilding an animal or a species. Most Romanian conservationists and locals understand it rather in this latter way. Hence, from this reintroduction perspective, European bison was brought to Romania as part of a species recovery programme, focused on the survival and growth of a wildlife population. For the people on the ground, the rewilding rhetoric of non-intervention and autonomy held meaning to be sure, and gave the general philosophy of how bison management was to be conducted. Yet, in practice, it was more complex. The reintroduction team walked a fine line, perpetually considering and reconsidering when and how to support or let loose, because they had the wellbeing of the animals at heart.

The composition of the bison founder group in the South-Western Carpathians was planned based on behavioural assumptions related to age and sex.³⁷ The imported animals were supposed to be at the right age for high reproductive potential and old enough to cope with new stressful situations. Bison reach maturity early, at two years of age, and they live on average fifteen to twenty years. The first imported bison were also picked to mimic a herd structure. Hierarchical and gregarious animals, they live in unstable matriarchies and have a polygynous mating system. One adult bull usually orbits in and around the female-dominated herd, and bachelor groups disperse around. Yet, becoming a herd and forming relations was not straightforward; the animals were strangers to each other; they did not share a culture of learned behaviours and relations,³⁸ and this was stressful, giving rise to aggression in order to establish dominance.

The first group had eleven young females, two older females and four young bulls. Romanița 4, a bison female from a nearby Romanian reserve was selected to be the matriarch. A ‘local’, she was assumed to be well adapted, and to arrive less stressed because of the shorter transport. Birk, a large adult bull from Kolmarden (Sweden) was the dominant male. Ulysses was the second adult bull, donated by Parco Natura Viva (Italy). He was thought to be a ‘fine specimen’. But the breeders feared he would not survive the transport by himself, and donated his brother Enea too, as a comforting partner, although Enea was not yet one year old, too young to survive reintroduction. He would be a ‘sacrifice’ for the greater good of the population.

This was a new experience not only for bison, but also for the humans involved. The Romanian team, composed of two WWF managers³⁹ and two young local rangers (henceforth I will call the four of them ‘the team’), had never worked with bison before. Hired consultants and a local veterinarian, experienced in rewilding and ungulates, aided them. All of them walked on new, uncertain grounds.

The life of the first herd would turn out to be hard and full of unplanned events. The animals were initially released in a fenced acclimatisation enclosure where they were fed and monitored (with a minimum of actual contact to avoid further habituation). The fifteen-hectare enclosure was located in the foothills, among beech trees, close to a forest road, three kilometres away from the nearly deserted hamlet of Plopu (belonging to the administrative commune of Armeniș). After a month in the pen, the animals were released into a larger area (163 hectares), enclosed with an electric fence. This enclosure was intended to last for one year, but in fact it ended up being longer, as bureaucratic approvals for the final release proved hard to achieve. Soon after arrival, one of the young females jumped the fence and took to the village. Used to humans from her previous captive life, she lounged in the sun and ate people’s hay and apples. The team left her be. They did not try to bring her back, as she seemed to be calm and locals did not mind her.

In the enclosure, in the first weeks, Romanița seemed to take on her assigned role of matriarch. As a leader, apparently she coordinated the movements of the group, and resting times. After release into the larger forest, she would drive the foraging and the scouting over the forest territory. Yet, another elder female, Isa (from Avesta reserve in Sweden), fought Romanița for leadership. Isa became the leader. The team saw Romanița giving up on fighting and only understood why this had happened a month after the reintroduction, when she gave birth to a calf.⁴⁰ It was a total surprise for the team that she was pregnant. Seemingly preganancies are not so visible, and even in reserves the females are not approached to check. So, they thought, she lost the leadership because she did not want to endanger the baby in her womb. A move for safety. At this point, it became clear that the ground team was in for many surprises; they cared for the animals and wanted to ensure their wellbeing, but many things were out of their control.

Romanița’s calf, Prâslea (Rom. The Little One), was born in June; the first European bison calf in Southwestern Romania. He became a little local celebrity, the newspapers acclaimed his birth and people voted on Facebook for his name. But the celebrations were not long-lived. A month later, the team found him dead, and the headlines read, ‘Tragedy: one of the most beloved symbols of our area died unexpectedly’. He was found with his jaw and ribs broken, signs that he was kicked. ‘This kind of accident happens in nature, and probably the bull hit the little one by accident while sparring with others’, said the press office, optimistic about the good nature of the animals.⁴¹ More plausibly, an adult bison male kicked him on purpose. Male infanticide is common in mammals where a few males monopolise reproduction, and scientists explain it in terms of the males trying to increase their mating opportunities by shortening postpartum infertility in the victim’s mother.⁴² Although they were initially open to media accolades around charismatic individuals and events, incidents such as Prâslea’s death cautioned the team against publicity and premature celebrations.

Hierarchies are important for the gregarious bison, and shifts often occur within herds,⁴³ so aggression occurs periodically. Following the death of her calf, Romanița grew weak. She was marginalised by the herd; pushed away from the feeding trays. In the beginning she made no effort to contest, but soon she regained strength and was competing to be matriarch again, as the managers observed by tiptoeing around in the bushes. In September, five months into the reintroduction programme, the herd was released in the larger rewilding enclosure of 150 hectares, and Isa led the herd. But her leadership was also short-lived. A week after the release, Isa died. Her necropsy report showed she had bluetongue disease, an insect-borne viral disease that spread through South-eastern Europe in 2014, killing domestic livestock; it was caught by bison probably from the sheep stationed in the area, or via the hay feed. Bluetongue affected the respiratory ways, causing fever and inflammation. Four of the bison that died in the first years were confirmed bluetongue carriers.

Meanwhile, Romanița got weaker and hung out closer to villages. The rangers worried for her and thought some form of intervention was necessary. They tried to understand what the problem was, and called on veterinary help. The vets gave her antibiotics and eventually in two weeks she was well again.

The beginnings were not encouraging. In retrospect, the team complained that the selection of this first group was not well-planned. They suspected the consultant rounded up available animals, who were not necessarily ‘fit’ for the wild. ‘It was a collection, not a selection’, the managers told me. For the next imported groups of bison, they fired the consultant and tried to do the selection themselves. They searched for animals in reserves where management was less heavy-handed, and animals were let to roam free, without much supplementary feeding. They negotiated with the reserves in advance, and tried to bring most animals from the same place, where they would be prepared for reintroduction.

The hope of the Romanian team was that, once released, the bison would ‘reactivate their wild genes’, start foraging, disperse into the forest. But this did not happen in the first years of the reintroduction. The bison seemed to have another plan for survival. They stayed put, close to the hay feeding areas. Not bothered by human presence, they approached at the rattle of feeding buckets. They resembled tame furry cows, showing clear signs of habituation and dependence. The bison were lethargic, the calves died, bluetongue struck.

Besides dying from diseases, the reintroduced bulls died from fighting each other, or from exclusion by others. The feeding from a point source seems to have enhanced exclusions. The managers explained that, as long as feed was artificially provided in one place, instead of being dispersed in the territory, it was possible for the older animals to exclude the younger ones to the point of starvation. Enea, Ulyssee’s young brother, whom the team thought from the beginning to be a sacrifice, did indeed die of starvation and isolation, because, being the youngest and the weakest, he was pushed away.

In the case of the other young bull, Zwitcher, exclusion was even more apparent. He was three years old, born in a reserve in Germany, where he was kept in an enclosure with no natural grazing areas. As this reserve reached its full capacity, the young ‘surplus’ bull was moved to a reserve in Belgium, from where he came to Romania.

The team confronted loss every few months and were terrified. They felt responsible for the death of the animals and cried over them. In an attempt to feel more detached, they stopped naming the newborns. The survival rates were not encouraging. Calves died. Feeding from a point source seemed to affect social relationships and to enhance the possibilities of marginalising individuals to the point of starvation. The team struggled to understand what was happening, as they tried to keep a respectful distance from the animals. A project with high expectations looked like it was going to fail. Something was obviously wrong. The team thought the management after release needed to change, to become tougher. But where to look for good practice? What had worked in the past? In the first half of the twentieth century, bison numbers had increased by careful husbandry in Yellowstone (US) and Białowieża (Poland), the two iconic places for bison rescue. Bison were handled like livestock. In Europe, supplementary feeding and selective culling had been for decades the policy recommended by the IUCN Bison Specialist Group through the conservation action plan, led by Polish breeding specialists.⁴⁴ Animals of the first imported group released in the South-Western Carpathians were seeking humans. They were not moving around to explore the territory. Some jumped the fence and took to the village. Thus, heavy handling that would increase dependence on humans was not the solution the Romanian team was looking for.

Two years into the project (2016), the team decided on a switch in management towards a hands-off approach, more in line with the non-interventionist approach of rewilding. ‘We said “no”. we have to let natural processes happen’ [Interview with Managers, December 2019]. They fired the consultants who advised hands-on methods, and opted for a hardline approach that would cause disruption, trusting the agency of bison to fend for themselves. As Tănăsescu aptly remarked, incorporating the animals’ responses challenged and modified the practice, towards a more relational understanding of the reintroduction project, in which the team trusted the wisents ‘to be able to learn, on their own, how to live in a new environment’.⁴⁵ The new releasees were only given one month to acclimatise – a ‘hard release’ that would eventually reduce reliance on humans for food. The team expected this abrupt change would inflict a high death toll, but thought that the unfit animals would die anyway sooner or later. So a high rate of short-term survival would not contribute to the flourishing of the population in the long run. Evolutionary ideas provided an avenue of understanding what might work as a long-term plan – concepts such as natural selection and the survival of the fittest: ‘If you do not let natural selection happen, you would not have strong specimens out there. If we fed them permanently, we would have habituated, defective animals.’ [Interview with Managers, December 2019]

This change, although it entailed risks, seems to have worked for the best: two healthy calves were born and survived, there were no more dead bison, and animals stayed away from the handling pens, and hiked for forage above 1,600 metres in altitude.

Tensions between degrees and forms of care are always there in reintroductions. How much care to give to animals bred with a view to reintroduction, or already reintroduced? How much food to supply, how much shelter and veterinary care? When would too much care hamper the development of skills, and when would a reduction of care kill the animals? And is the death of animals something to be prevented, or something natural, a regulatory mechanism beneficial in the long term? If the animals die of starvation, is this natural selection, or a tragic loss? These decisions never cease to challenge people involved in reintroductions. The tensions between care and intervention are deeply embedded in norms of social legitimacy, in foggy notions of wildness, habituation and proper animal behaviour.

Back in the 1950s, the first reintroduction of European bison to Poland in the Białowieża forest practised heavily interventionistic management, modelled on animal breeding practices. This hands-on approach continued throughout the decades until very recently, being the type of management advised by the wisent conservation experts related to IUCN.⁴⁶ In the first five years of the Polish reintroduction process (1952 onwards), only two bison were released every year, and some of the animals were removed (placed back in the enclosure), as it became obvious that they were seen as reproductively or behaviourally inappropriate – sterile, too aggressive, too adventurous (they crossed the borders of the Białowieża forest).⁴⁷ In twenty years, by the end of the 1960s, from 38 reintroduced bison, the population size increased nearly sixfold, reaching 141 bison.⁴⁸ With the exception of the first two years (1952–1954), the wisent were fed regularly. For up to six months every winter, large congregations of bison hung around the feeding areas. Heavy management continued because of fear of conflict with foresters and, most notably, with farmers – if hungry bison were to debark trees, damaging timber production, or to raid haystacks and crops.⁴⁹ The optimal number of wisents was subject to political negotiations, and hence to regulation, as, from the 1970s, the bison started to be perceived as an overabundant species, while continuing to be hailed as a symbol of the nation and of the struggle for conservation.⁵⁰ From 1971, culling started as a management method. In the next four decades, nearly 900 bison were culled, a few because they were blind, some for entering farmers’ fields, and large numbers because they suffered from painful diseases⁵¹ – belanopostitis, known as pizzle rot, tuberculosis and pulmonary diseases.

Over the decades, it has remained controversial whether a population that survives and reproduces by being fed and culled represents a real wildlife conservation success. Moreover, feeding by humans certainly points at the volatile and fuzzy character of the term ‘wild’,⁵² which had been used in wisent-related discourse. In the 1960s, the relation between human intervention and wildness was also at the core of heated debates about the reintroduced wisent that occurred in the Polish professional community of zoo specialists, veterinarians, breeding experts, foresters and wildlife biologists. Jan Zabinski, the president of the International Society for the Preservation of the European Bison in the postwar era, renowned zoologist and zoo director, considered that feeding distorts the morphology and mentality of the wild wisent turning it towards domestication,

Despite these influential voices, proponents of heavy management dominated debates and practice for a long time. They aligned with the desires of trophy hunters and foresters. The breeder-managers understood the situation in the framework of complex social relations and networks of institutions operating around the Białowieża forest,⁵⁴ which lent legitimacy to interventionist practices. The wisent was perceived as a fragile and precious ‘king of the forest’, in need of sustained human help and regulation. A prevalent fear existed that a growing wisent population would trespass into inhabited areas if the feeding stopped. Studies in the 2000s have shown that ithe growing population of wisent did indeed often migrate into agricultural areas and depredate crops, causing conflicts with locals and the necessity of monetary compensation.⁵⁵

Yet, from 1990 onwards, the influence of non-interventionist ideas began to grow. Wisents were increasingly seen as part of the ecosystem, animals able to migrate freely, regulated by natural selection.⁵⁶ Despite this growing trend, by 2007, the Polish government adopted a wisent management plan that enforced a paradigm focused on heavy handling and intervention.⁵⁷ Polish wildlife biologists would contest such practices relentlessly, and only in the 2010s would their plea gain traction.⁵⁸

‘To feed or not to feed, that is the main question’, said Tomasz Kaminski, a biologist at the Mammal Research Institute at the Polish Academy of Sciences, to the Guardian in 2011.⁵⁹ ‘The bison’, says Kaminski, ‘are semi-natural but on the brink of being able to fend entirely for themselves. They live freely in summer. But in winter they are fed by rangers, taming the beasts.’ ‘By intensive supplementary feeding, [national park] managers increase bison survival and increase their reproduction’, said Rafal Kowalczyk, the head of the same Institute.⁶⁰ But this also increases the population beyond the ‘carrying capacity’ of the forest, and ‘Then they cull bison because there are too many individuals’.⁶¹ These scientists, in disagreement with the longstanding dominant policy of wisent management in Białowieża, argued that the feeding should be phased out and the cull should stop, leaving death to claim its victims naturally because ‘Nature is a better selector than humans’ in pruning weak animals out of the herd. Experts showed that, inherited from nineteenth-century management practices, the feeding had a negative impact on wisent health, overall making their physical condition worse,⁶² reducing movement and increasing the load of parasites.⁶³ All in all, it is considered that management based on zootechnical practice, inspired by livestock farming, slowed down the ability of European bison to naturally adapt.⁶⁴ Under the pressure of scientific results, both wisent management in Poland and the general guidelines as to how this species should be managed are shifting in recent years away from a breeding-controlling perspective and towards a rewilding-autonomy perspective.

Most reintroduction projects face uneasy coexistences with local populations, this usually being part of the reason why these species had dwindled towards the edge of extinction in the first place. In the Polish case just described, the fear of wisent–human conflict kept control and interventions at high levels. In the case of contemporary Romania, I will show that fears of local wisent–human coexistence gone wrong also made certain interventions seem necessary, although the undergirding philosophy of this project emphasised that the autonomy and agency of wisent should develop freely.

In 2019, I went to Armeniș for the Christmas holidays with my partner Arryn. It was mostly still and it smelled like smoke. Houses with red terracotta-tiled roofs sparsely dotted the eroded hills covered in scrub. The undulating slopes in the late December landscape were all shades of browns and greys: reddish clays, burnt amber grasses, patches of dark brown ploughed earth, greenish brown shrubs bared of leaves (Figure 2). In places, the curves of the landscape sank into hollows: damp valleys carved by creeks in purplish stone. Grim valleys. The weather was mild, giving signs of an easy winter for the bison. The winter before had been hard, extreme. The ground had been frozen solid for days. The eight new bison ‘imported’ that year had died. The rangers braced for more deaths in 2019. They were committed not to give much supplementary feed, and fortunately there was no need, as 2019 was a mild winter for the bison.

The day after Christmas, the rangers took us to check the bison territory. Up in the forest, there was snow; a slushy snow, melting under the wheels of the jeep. The rangers had put one bale of hay and a mineral salt lick down near the road, in case the bison were starving – a form of intervention that still maintained distance. The provisions were untouched so far.

As we drove up, another jeep came down. On the narrow road, as the cars squeezed past each other, both stopped, and the drivers lowered the windows. In the other car were two men in camouflage gear with rifles. The driver greeted us gruffly. He had fair complexion, turquoise eyes and a teasing grin. ‘We saw your bulls up on the other side, close to Rusca’s area’, ‘How many?’, asked one of the rangers, ‘Thirty or so’, said the driver after a pause, and laughed. The rangers laughed, complicit, saluted, and closed the window. ‘Thirty bulls’ was a joke about the exaggerated reports often given by shepherds. Bison never stay together in such large herds (on average a herd has ten bison). When the driver said ‘your bulls’ (boii voștri), it sounded like saying they were not truly wild bison, zimbri, just some trivial domestic bulls. But perhaps I was wrong. Wild bulls (boi sălbatici) was also an ancient name for male bison in Romania.⁶⁵ The camouflage-gear men were hunters. They were patrolling the forest territory tracking game and checking on the feeders they had put up as a game management device. The brief jeep encounter gave away a tension between the rangers and the hunters – I felt it in the gaze of the men, measuring each other, in the smiles, the pauses, the teasing. The driver used to be the head of the hunting association. After many years of being the boss, earlier in 2019 he was demoted, as he was allegedly caught poaching in heart of the neighbouring national park, Semenic Mountains, but the court case was mysteriously closed after two years of hearings.

The wisent is officially considered ‘game’, vânat, that cannot be hunted. This categorisation (perhaps existent due to bureaucratic rigidity, lack of imagination and will for change) means the animals are somehow under the jurisdiction of forestry districts and hunting associations. Romanian hunters historically contributed to species conservation, as during socialism they were employees of forestry districts, in charge of managing wildlife. Yet, this management followed an economic-productivist logic, in which annual ‘quotas’ of game were ‘harvested’ sustainably. The hunters we encountered that day, like nearly all hunters in the area – about sixty were active around Armeniș – were mildly hostile to bison reintroduction, and to conservation (as practised by conservation NGOs) more generally. Bison monitoring encroached on their territory. As the wisent managers explained, ‘We intruded where previously the hunters did as they pleased’; the hunters complained that bison interfere with other species, with the feed provided for managing other game, like boar and deer. The hunters compared the video camera surveillance meant for bison to the means of coercion and repression of former socialist dictatorial times. The relation between reintroduced wisents, rangers and hunters was an uneasy one, which complicated the promoted policy of non-intervention.

***

Unlike the hunters, the majority of the local population of Armeniș was very supportive of the reintroduction, which I noticed as my involvement with the project began in 2016. I was hired as a consultant for WWF on the local acceptance of bison reintroduction, and I conducted an extensive survey, entering people’s houses with questions, and showing them pictures of bison to elicit their opinions (figure 3).⁶⁶ Most of them were delighted with the project. Their eyes lit up at the sight of my bison pictures and, in a profusion of emotion, one man even kissed the photos. This acceptance was still there when I repeated the survey in 2020.

People here seemed better-off than other rural inhabitants of Romania. They were middle class rural gentry, with an outstanding appetite for celebrations – village events with expensive musicians and glamorous glittery outfits, a rural opulence that felt like an ebbing quiver of former glory. This area had been one of the most industrially developed in the country. People in Armeniș used to work in jobs that disappeared after the collapse of socialism, and left local youth with the only option of emigration abroad. My studies in 2016 and 2020⁶⁷ found an overwhelming acceptance of what the locals perceived as an awe-inspiring, beautiful and gentle animal, with the exception of the few hostile hunters and a few people whose fruit-trees were ‘damaged’ by bison. People thought that bison were not so ‘wild’, resembling furry cows. Informed by a history of caring for livestock, they worried about bison surviving winter on their own, they thought they were vulnerable and should be taken care of.

Economic gain was expected from the project, as locals were fed an insistent tourism-based development narrative: bison would make them money, by providing jobs and tourists. This market-and-development conservation discourse was championed by Rewilding Europe, but five years into the project, locals felt many promises had not been met. Yet, for many of the people, these promises were not the reason for caring about the project. They were enthusiastic about having bison in the area, welcoming the initiative to improve the lives of vulnerable animals. They understood the perspective of saving a species from extinction. As Dolly Jørgensen has argued, recovering lost species is an emotionally-charged practice, and sentiments of loss matter deeply and motivate people⁶⁸ to care about the bison’s future. Yet, most people upheld ‘separationist’ views (bison belong to the forest), and adhered to the logics of wildlife control rather than to those of non-intervention.

Coexistences are uneasy. Nearly every other history of bison conservation, both in Europe and in the US, has involved conflicts with local residents, because of migratory movements into farmland, raiding of crops and haystacks, seen as transgression, threats to existing hierarchical relations between humans, livestock and bison.⁶⁹

In 2017, a female bison came down to the village in Armeniș. She was skinny and wretched. In the case of the big groups of Yellowstone and Białowieża, migration was usually in herds and understood as something undesired but natural for a free-ranging population. But here the movement was seen as an individual ‘disorder’, a condition of the out-of-place former captive. The managers suspected that, because this female bison came from a zoo, she was unfit, maladapted to live in the forest, too much used to shelter, feed and people. Most people were surprised to see her in the village, but did not really mind her. Yet, the vice-mayor, commenting for the press, expressed a well-worn cliché of undesired otherness: ‘The bison does no harm … but people panic because it is a wild animal after all … Clearly, its place is not in the village.’⁷⁰ The rangers attempted to chase her away back to the forest, although they saw it was a hopeless case, ‘She refused to advance, probably because she sniffed traces of bear or wolf. After two days, she was back in the village. We chased her again.’ (Interview for press 2017). Later, she lost weight and died. The managers thought human help would not have prevented her death.

But it wasn’t only ‘misfits’ from zoos that came down to the village. The next year, 2018, it was the largest male of the herd who came. Birk weighed nearly a ton. He was reintroduced from a Swedish reserve, where he was not heavily handled. Birk in the village looked like huge trouble. He chomped at apple trees and hung out near the school. Kids approached him to take selfies. The team feared he would become aggressive. The rangers chased him away, to the point of exhaustion, but to no avail. The managers were against immobilising Birk to transport him back to forest, but the rangers pressed for immobilisation as they saw no results from chasing. Moreover, they thought that chasing meant killing, as ‘They are already in weak condition when they come down’ [Rangers interview, December 2019]. After two months, they decided for immobilisation, despite the risks and stresses of such an action, ‘We had to move him back up in the mountains, for Birk to understand it was not ok what he did’, said the managers. But to tranquillise an animal protected by law, the team needed bureaucratic approvals, which did not come easily. This situation was unprecedented in Romania – the legal status of the bison was not clear, so no bureaucrat wanted to get anywhere near the case. Although it was presented it as a great emergency, the papers took a month and a half to arrive.

The team gave agency to Birk. He had to learn the terms, not to have his will simply crushed. But he needed to be disciplined and to understand the boundaries of his coexistence with humans; he belonged to the forest. Certainly there was benevolence in the conservationists’ actions but, ultimately, enforcing the boundaries necessitated the use of forceful intervention. This represents a part of the complex arrested autonomy that species to be saved from extinction sometimes face.⁷¹ And perhaps the bureaucrats’ inaction inadvertently facilitated imagining the possibility of coexistence. For six weeks the largest bison bull of the herd lived on the village streets and it did not end in catastrophe.

The movement of wildlife means encounters, and it is thoroughly infused with contingency.⁷² The process of bison-becoming-wild entails uncertainty. Human–bison encounters are seen as clashes. They move beyond boundaries, threatening anthropocentric structures in place. From the conservationists’ vantage point, intervention is required to maintain spatial structures of separation. The flux of migration typical of the species must be moulded or eliminated. The ability of these non-humans to inhabit landscapes in fully autonomous and diverse ways is not yet acknowledged. But can human–bison coexistence without intervention and control, with both species living as they please, even be remotely imagined as happening in present-day Europe?

In species conservation, non-intervention is complicated in practice, despite being commended as a way of fostering non-human autonomy and agency. Conservationists involved in recovering species from extinction walk a fine and messy line, in which the threads of care and detachment are interwoven.

When European bison recovery was initiated in 1923, conservationists knew very little about what the rescue of this species would involve. They knew that they needed to ‘make more’ bison, as their numbers were down to a daunting few, the species having become extinct in the wild in the midst of wars. Making more meant breeding them. In the following decades, the reintroduced bison in the Polish Białowieża Forest, the stronghold of the species, were managed heavily. The Polish experts intervened strongly in what they called ‘living free’,⁷³ guided by a ‘breeding’ philosophy. Bison were thoroughly taken care of, moved around, fed, culled, freeze-branded – far too much, according to several experts involved, which led to dependency and unfitness. In a realm of controversy, policies of intervention did not stop once the wisent population grew. Amidst disputes over forest use, more intervention and control were deemed necessary to keep the wisent from ‘damaging’ crops and forests. A philosophy of care and control usually has public appeal. Therefore, authorities in charge of managing an animal perceived by the public as an important national symbol walked on treacherous terrain. They did not want to risk a policy of non-intervention that would attract the critique of carelessness and cruelty (starving the animals, allowing them to suffer). A pathway was laid for a complicated human-wisent relationship.

Nearly sixty years after the first Polish release, the South-Western Romania reintroduction project was very different. First, the context was different. Here, the bison recalled local legends but were not as symbolically charged and prominent as in Poland. Therefore, the project began as low profile, sheltered from public eyes and egos. This meant that unfortunate bison events, like deaths or disease, did not cause much public commotion, and the rewilders could follow their plans. Moreover, the project began at a time when the wisent population was considered safe; therefore a failure would not have endangered the entire species. Also, this project took place when the philosophy and discourse of rewilding (which in essence emphasised non-intervention as opposed to the more interventionistic policies of reintroduction) were well underway in conservation. Non-intervention, and its corollaries – ‘nature knows best’, ‘natural selection is the best manager’ – provided guiding principles. The predicament of this approach is to recognise the autonomy and agency of wildlife, and of nature more broadly. The Romanian team championed a hands-off approach, not necessarily following contemporary trends in conservation (although these provided a baseline reassurance), but rather in response to the agency of the bison themselves, who behaved poorly under hands-on management, with low rates of survival, high rates of fighting and habituated behaviour. What Tănăsescu called a ‘loosening of control that resulted in new relations between people, animals, and territory’⁷⁴ had apparently worked well so far.

However, even in these circumstances favourable to non-intervention, it proved difficult to keep up barriers between wisent and people. In the end, it proved very difficult to relinquish control, and to trust that the animals do what is best for them.⁷⁵ Intervention has been necessary, in order to move the large mammals away from villages. The hands-off approach in the South-Western Carpathians was challenged by separationist views held by local people fearing ‘damage’, whose notions of what wildlife is, where it belongs and how to co-exist are embedded within historical relationships. In this view, humans must intervene to return wisent to their rightful place in the forest. At this point, the two projects were somewhat similar: the coexistence between wisent and farmers requires management, and the farmers’ wellbeing has priority, historically and in the present, in Poland as well as in Romania.

In the case of such trespassing events, where bison descended to inhabited areas in search for food, the principles of autonomy and agency came under scrutiny. The team believed that the short-term autonomy of the animals (to do as they please) would create a pathway of habituation, which would endanger the long-term autonomy – understood as a separation between humans and wisent. The contemporary endeavours of reintroduction as a radical species conservation tool, legitimated by alarming rates of loss, do not envisage a becoming-with humans, under control, as breeders did for the first half of the history of saving European bison.

Contemporary conservationists envisage reintroductions as a becoming-apart, a becoming free, a future that stores emergent qualities and unforeseeable natural processes.⁷⁶ But the predicament of non-intervention is complicated. Salutary as it may sound, non-intervention is limited in practice, because it grafts onto relational circumstances, onto contentious coexistences, and a history of control and dependence, which have changed wildlife, and profoundly transformed the animals into natureculture,⁷⁷ beings that are inextricably a part of both human and nonhuman worlds.

In the long run, the struggle to save the wisent from the brink of extinction has turned very slowly from ‘hands-on’ to mostly ‘hands-off’. But a constant short-term weighing of various forms of intervention is still common practice, and it will perhaps remain so for a long while. The relationships between humans and wisents are considered on a case-by-case basis, each situation entailing new dilemmas, and new considerations about ‘how close is too close?’ and ‘how far is too far?’ Practices of recovering the free lifeways of endangered species were and still are ridden with profound uncertainties and ambivalences, contingent upon particular conjunctures. Liberations are messy in practice and autonomy is a matter of degree. Caring for non-human life involves limits and hazards.

During the writing of the article, the author was funded as part of the NWO Vici project VI.C.181.010, “Moving Animals: A History of Science, Media and Policy in the Twentieth Century”. The author is grateful to the first managers of the wisent reintroduction project, Alexandru Bulacu and Adrian Hăgătiș, who gave her the opportunity to work on the topic, and openly discussed their experiences and challenges. She feels equally indebted to the local rangers and people who shared their work and time during fieldwork. Many thanks are due to the two anonymous reviewers, as well as to Stefan Voicu, and to her partner Arryn for insightful feedback.